Lens Cell Populations Studied in Human Donor Capsular Bags with Implanted Intraocular Lenses

Abstract

purpose. Posterior capsule opacification is an ongoing cellular redistribution process. The level of viable cell coverage was therefore determined in human donor capsular bags with implanted intraocular lenses, and cellular morphology and ultrastructure were investigated in relation to cell type and level of differentiation.

methods. Donor capsular bags, retrieved at intervals of 4 months to 13 years after surgery, were investigated by phase optics before fixation. Postfixation techniques included scanning electron microscopy and transmission electron microscopy of sections and immunofluorescent staining of cytoskeletal proteins in wholemounts.

results. All the capsular bags contained a large population of viable cells on the capsular surfaces. Cells on the anterior face of the anterior capsule and in the spaces around the intraocular lens had an elongated morphology and expressed α-smooth muscle actin. The cells formed light-scattering, multilayered aggregates and strands that were surrounded by layers of extracellular matrix. The regions between the intraocular lens and the equator of the bags were populated by monolayers of epithelial cells of normal morphology and ultrastructure, on both the anterior and posterior capsules. In some regions the apical surfaces of the two epithelial monolayers were in contact, and in some parts of the equatorial regions, differentiation of cells into well-organized fiberlike cells was evident.

conclusions. Human capsular bags contain a large population of viable cells for many years after cataract surgery. Cells in the regions around the intraocular lens undergo transition to a mesenchymal type. Cells peripheral to these regions can form a stable closed microenvironment in which both normal epithelial morphology and differentiation to fiberlike cells are maintained.

Posterior capsule opacification (PCO) causes a secondary loss of visual acuity in approximately 30% of patients (in almost 100% of young patients ¹ ) within 3 to 5 years of cataract surgery. ^{2 3} Not only does this have psychological consequences for the patients, but the necessary secondary intervention by Nd-YAG laser can itself result in further complications ^{3 4} and increase the financial consequences of cataract surgery. The problems associated with PCO also impede the introduction of intraocular lens (IOL)–based techniques in the developing world, where the need for them is greatest. 

Investigations into the nature and development of PCO have involved clinical studies (reviewed by Apple et al. ³ ), experiments with in vivo animal models ^{5 6} and experiments with animal and human lens epithelial cells (LECs) in tissue culture. ^{7 8} Clinical investigations by specular microscopy have shown the presence of cells on the anterior IOL in the immediate postoperative period ^{9 10 11} but cell numbers decline to very low levels after 3 months. ¹²  

PCO results from the resilient growth onto the posterior capsule (PC) of residual lens epithelial cells. ^{3 4} The first clinical appearance of cells is that of a cellular sheet growing across the capsule, ^{13 14} but this cell monolayer seems to have few visual consequences. ^{3 15} Regression of this cell monolayer occurs in some patients, leaving a clear and transparent central PC. ¹⁶ In other cases cell regression from some areas of the PC, multilayered cell aggregate formation, increased deposition of extracellular matrix (ECM), and wrinkling of the PC occurs. ¹⁷ These are the changes associated with light scatter and loss of visual acuity. In addition, fiberlike development of equatorial cells can lead to the formation of a peripheral Soemmerring’s ring and cell swelling to the formation of Elschnig’s pearls, which can occlude the visual axis. ^{3 18 19}  

Cells with a fibroblastic morphology have been observed in both clinical and experimental studies of PCO. ^{3 20 21} Immunohistochemical techniques have detected the presence of theα -smooth muscle isoform of actin (α-sma) and several isoforms of collagen in sections of human capsular bags. ^{22 23} Because normal LECs express only F-actin and produce ECM containing mainly collagen type IV, ²⁴ it seems that some PCO cells undergo transition to another cell type. Epithelial-mesenchymal transition is known to occur in ocular tissues and tissue-cultured cells. ^{25 26 27 28} Posttransition cells are characterized by fibroblastic morphology, α-sma expression, and increased deposition of ECM containing multiple isoforms of collagen. 

Although the pathologic course of clinical PCO has been well described (reviewed by Apple et al. ³ ) the cellular changes, and the mechanisms underlying them, have not been fully elucidated. Peroxidase-based immunohistochemistry of sectioned specimens provides little information about cell structure or viability or the relationships between cells. An understanding of cell growth and transition in all regions of the capsular bag and of the driving forces involved is essential for research into possible blocking mechanisms. Recent in vitro experiments with a human capsular bag model ^{29 30 31} have shown that residual LECs have the capacity to proliferate rapidly on the lens capsule in defined media, even in the absence of added serum or growth factors. Epithelial-like cells populate all the capsular surfaces and during prolonged culture light-scattering wrinkles and cell aggregates form, mimicking the in vivo situation. This model is currently being used for kinetic studies of PCO and potential inhibitors. 

The data obtained in the present investigation of in vivo capsular bags provide a baseline of information on cell distribution and cytoskeletal ultrastructure against which cell growth and differentiation in human capsular bag models can be assessed. 

Methods

Capsules were obtained from human eyes donated for corneal transplantation from the East Anglian Eye Bank (Norwich, UK) and the Cornea Bank (Amsterdam, the Netherlands), or after death from the Department of Ophthalmology, University of Liege (Belgium). The research followed the tenets of the Declaration of Helsinki regarding the use of human material. 

Immunocytochemistry

Twenty-one capsular bags with IOLs were obtained for this study. Donor ages ranged from 57 to 87 years, and the time elapsed from cataract surgery from 4 months to 13 years. Controls were capsule–epithelium preparations from 14 donor eyes, in the age range 6 to 76 years. 

East Anglian Eye Bank capsular bags were dissected from the globes and placed in Eagle’s minimum essential medium (EMEM). Structural features from anterior and posterior aspects were recorded photographically by dark-field and phase optics. Specimens were fixed in 4% formaldehyde in phosphate-buffered saline (PBS). In most cases, after 15 minutes’ fixation, the IOLs were removed to allow detailed observation of cells on the capsule. Fixation was then continued for at least 1 hour. IOLs were stained separately and mounted in cavity slides for observation. 

Prefixation was applied to the control capsule–epithelium preparations by introducing fixative into the globe. After 1 hour, the lens was dissected and placed anterior face down in a petri dish. The PC was split in a star configuration, the flaps were pinned out, and the lens body was carefully lifted away from the epithelium. The preparation was covered with Ca²⁺-free PBS while residual lens fibers were removed with microforceps and was then returned to fixative. This method ensured that the epithelial cells were fixed before any changes in structure could be caused by the dissection process. 

Preparations were permeabilized in PBS (pH 7.3) containing 0.5% Triton X-100 for 30 minutes. Nonspecific sites were blocked with appropriate serum (1:50 in 1% bovine sermum albumin [BSA]/PBS). Anti-α-sma (Clone 1A4) was applied overnight at 4°C, anti-vimentin (Clone V9) and MIP26 antibody (a gift from Joseph Horwitz, Jules Stein Eye Institute, University of California at Los Angeles) for 60 minutes at 35°C. Visualization was with fluorescein isothiocyanate–conjugated secondary antibodies for 60 minutes at 35°C. The F-actin cytoskeleton was stained with Texas red-X-phalloidin (Molecular Probes, Leiden, The Netherlands) for 30 minutes, and cell nuclei with DAPI for 15 minutes, both at room temperature. The stained and washed preparations were mounted in medium (Vectashield; Vector, Peterborough, UK). Specimens were viewed with a microscope (Standard R; Zeiss, Oberkochen, Germany) and images recorded on film (Ektachrome 400 or Tmax 400pro; Eastman Kodak, Rochester, NY), and with a confocal microscope with cooled charge-coupled device (CCD) camera (Viewscan DVC-250; Bio-Rad, Princeton Instruments, Marlow, UK) and software (MetaMorph; Universal Imaging, West Chester, PA). After initial observation, the capsular bags were unmounted and washed and the layers further separated to allow observation of the anterior and posterior capsule individually. 

Electron Microscopy

Sixteen capsular bags were prepared for electron microscopy. Donor age ranged from 63 to 93 years, and time elapsed from cataract surgery ranged from 6 months to 4 years. After trepanation of the cornea, the eyes were immediately placed in fixative at room temperature (0.08 M cacodylate buffered 1.25% glutaraldehyde and 1% paraformaldehyde [ pH 7.3]). After fixation for several days the capsular bags were dissected and photographed using dark-field stereomicroscopy, and the IOL optic was carefully removed. The tissue was postfixed in OsO₄, and representative pieces were further treated for electron microscopy: for scanning electron microscopy (SEM), dehydrated in ethanol, dried with hexamethyldisilazone (Sigma–Aldrich, Zwijndrecht, The Netherlands), mounted, and coated with platinum/palladium, and viewed in a scanning electron microscope (model 505; Philips, Eindhoven, The Netherlands); and for transmission electron microscopy (TEM), dehydrated, and embedded in epoxy resin. Semithin sections were stained with toluidine blue and photographed. Ultrathin sections (∼80 nm) were contrasted with uranyl acetate and lead citrate and studied in a transmission electron microscope (model CM12; Philips). 

Results

General Appearance

The overall structure of a representative capsular bag is shown in Figure 1A , which is a montage of dark-field photomicrographs from the posterior aspect. A sectional diagram (Fig. 1B) demonstrates the corresponding regions of interest. In this specimen the IOL, which had been in situ for 2 years, was centered in the bag, and the rhexis was clearly visible through it. Fiberlike material in the equatorial region of the bag (region 1a, and arrows) appeared bright due to light scatter. The greatest variation between specimens was in the amount of this material. Representative PCOs (Figs. 1C 1D 1E 1F) show the variation from very little (Figs. 1A 1B 1C) to Soemmerring’s ring formation (Fig. 1F) in the equatorial region. Most specimens also had regions where the two capsular leaflets were in close apposition (Fig. 1A 1B , region 1). These appear as dark regions in the montage. 

Thickened, wrinkled (fibrotic) areas of capsule around the rhexis (Fig. 1A , region 2) and beneath the periphery of the optic (region 3) also scattered light. Again, the area of fibrotic capsule varied among specimens (Figs. 1C 1D 1E 1F) and was regionalized within individual bags. Islands of cells on the PC (Fig. 1A , regions 4 and 5) also scattered light from their margins but, in the specimens obtained for this study, did not extend to the central region of the capsule. Elschnig’s pearls (Fig. 1E) were found in only two of the bags. 

The combination of phase-contrast and EM observation, together with cytoskeletal staining, made it possible to assess cell coverage and morphology in all the capsular bags. The images presented are representative of the various regions of interest. 

Epithelial Morphology in Regions 1 and 1a

The structure of the cell layers on the anterior capsule (AC) and PC covering the region between the adhesion zone outside the IOL optic and the equator (Fig. 1B , region 1) was very similar in all the specimens, regardless of the length of time elapsed since surgery. Epithelial cells on the AC retained the cobblestone appearance of cells in a normal intact lens and had a similar cytoskeletal fine structure. Figures 2A and 2E show the F-actin in the apical borders of these regular cells in a capsular bag and a control epithelium, respectively. Fine details of the F-actin cytoskeleton were comparable and showed a dense apical mat with strong cell–cell contacts (Figs. 2B 2F) , a cortical submembranous layer in the lateral borders (Figs. 2C 2G) and whorls of short, stress fiber–like structures basally where the cells contacted the capsule (Figs. 2D 2H) . The vimentin cytoskeleton and microtubule arrays in the capsular bag AC cells were also comparable to those of normal epithelia (data not shown). 

The confluent monolayer of cells attached to the PC was seen most clearly in TEM images (Fig. 2I) . These cells had organelles and inclusions similar to the cuboid AC cells but were flatter, contained more vesicles, and showed more intercellular and subcellular spaces. The apical surfaces of the AC and PC cells were frequently in very close contact (Fig. 2I) , and the presence of two layers of nuclei in only slightly different planes of focus (Fig. 2J) emphasized this proximity. The nuclei were comparable in size and shape to those of normal lens epithelial cells (Fig. 2K) . No evidence of α-sma expression was detected in capsular bag epithelial cells or anterior cells from control epithelia (data not shown). 

Structure of Material Located in Region 1a

Variable amounts of fiberlike cells were found at the equator of the capsular bags (Figs. 1C 1D 1E 1F) . Where the amount was small, fiberlike cells, with homogeneous cytoplasm and very small intercellular spaces, were present between the monolayer of epithelial cells attached to the AC (Fig. 3A ) and the well-ordered lens bowlike cells on the PC (Fig. 3B) . Where a thicker region was present, the bowlike regions contained cells with elongated nuclei (Figs. 3C 3D) typical of lens fiber development and adjacent fiberlike cells with homogeneous cytoplasm and devoid of nuclei. Short runs of meridional rowlike cells were present in a few specimens. These can be seen (arrows) in Figure 3D in the plane of focus of the epithelium, before the nuclei become elongated in differentiating cells and pass into other planes of focus. The confocal images (Fig. 3E 3F) show the submembranous F-actin cytoskeleton of the anterior part of these regions. The cuboid anterior epithelial cells, the long axes of fiberlike cells, and the close apposition between the epithelial cells and the ends of the fiber cells appeared to resemble those of normal lenses. The membranes of the fiber cells contained interdigitations and gap junctions (Fig. 3G) similar to those in normal lens and showed a positive reaction to an antibody to the lens-fiber–specific membrane protein MIP26 (Fig. 3H) , even in the bag donated 13 years after IOL implantation. 

Regions Near the IOL

The physical presence of the IOL in the capsular bag resulted in a buildup of cell numbers and multilayering against the edges of the plastic. This was most easily observed (Fig. 4) at the equator against the haptic but occurred also against the optics. The F-actin staining of cells alongside the haptic (Fig. 4A) shows that multilayers of elongated cells with elongated nuclei (Fig. 4B) had formed arrays parallel to the haptic. These layers appeared very bright because of the close alignment of many F-actin–lined cell membranes. In the remainder of the epithelium, normal cell outlines and ovoid nuclei were present. 

Region 2: Rhexis and AC Overlying the IOL Optic

In all specimens, cells were observed on both the anterior and posterior faces of the AC in the region round the rhexis (Fig. 1 , region 2). Cells closest to the rhexis, on both surfaces, had an elongated morphology and formed long strands in concentric rings surrounding the rhexis. These cells had elongated nuclei (Fig. 5B ) and expressed both α-sma (Fig. 5A) and F-actin as a submembranous cortical layer. Cells on the anterior face of the AC extended out toward the equatorial region, and those remote from the rhexis formed confluent, epithelial-like sheets (Fig. 5C) . These cells expressed bothα -sma (Fig. 5D) and F-actin in stress fibers. All cells on the anterior face of the AC thus appeared to be α-sma positive. The multilayered network of cell strands surrounding the rhexis on the posterior face of the AC can be seen in a phase micrograph (Fig. 5E) . These cells also had an elongated morphology and were positive forα -sma (Fig. 5F) . 

Region 3: PC beneath the Optic Edge

Wrinkles in the PC beneath the optic were present in all the specimens (Figs. 1A 1B , region 3, and Figs. 6A , 7 A). As shown in Figure 6A the wrinkles formed circular parallel arrays at the margin of the optic or radiated from this region toward the central PC and caused considerable light scatter. The cells did not form confluent sheets but rather multilayered arrays of islands and ribbons with smooth edges that were devoid of cell extensions (Fig. 6B) . Many of these arrays were not intimately attached to the original capsule but were surrounded by large areas of ECM of loose composition (Fig. 6C) . Epithelial-like cells were rare, but two types ofα -sma–positive cells were common. Many cells were very elongated and myofibroblastic with elongated nuclei (Figs. 6D 6E 6F 6G) , whereas others showed an intermediate morphology, being hexagonal with ovoid nuclei but without prominent actin fibers. In both these cell types, F-actin and α-sma were present as submembranous cortical layers. 

The PC wrinkles themselves were frequently unpopulated by cells. Loops of elongated, α-sma–positive cells were present above capsular wrinkles that were devoid of cells (Fig. 7A) . Three possible stages in the development of this state are shown in Figures 7B 7C and 7D . Some deep wrinkles contained cells surrounded by loosely arranged, and presumably new, ECM (Figs. 7B 7C 7D) . Cell degeneration, as in one of the groups of cells in Figure 7C , could result in accumulation of cell debris within the new ECM that would remain in situ. In some cases the ECM was present as distinct layers close to the groups of cells. A deep wrinkle, without a cell population but filled with ECM and cell debris, is shown in Figure 7D . 

Regions 4 and 5: Central PC

The posterior capsules in the region of the optical axis were largely clear of cells and wrinkles in the specimens obtained in this study. Isolated patches of viable cells were present in the region between the central PC and the wrinkled area at the optic edge (Fig. 1 , regions 4 and 5). The cells formed multilayered colonies with smooth edges that scattered light (Figs. 8A 1A ), sometimes with an outer layer of epithelial-like cells (Fig. 8C) . Most of the cells were of intermediate morphology and expressedα -sma (Fig. 8B) , although the staining appeared to be cytoplasmic. A few large, substrate-attached epithelial-like cells with stress fibers were present (Fig. 8B) . 

IOLs

Stained IOLs were viewed from both surfaces, and it was possible to assess cell coverage over the whole surface at low magnification. Very few cells remained attached to the IOLs, and none were found in a central location of the optic on either surface in any of the specimens. Some adherent, isolated micronuclei, without cytoskeleton or cytoplasm, were present in most cases, indicating that a larger cell population had been present at an earlier stage. Macrophages, which have been observed on IOLs in the immediate postoperative period, ^{3 9 10 11 12 23} were not present on any of the IOLs in this study. 

Discussion

Previous studies of PCO (reviewed by Apple et al. ³ ) at macroscopic and light microscopic levels, and largely from a clinical viewpoint, have revealed few details of cellular structure. EM techniques ^{3 17 18 19} and immunoperoxidase localization of specific proteins in sections of specimens ^{22 23} have provided some details from selected regions. The combination of techniques used in the present study has provided an overall picture of cellular content in wholemount specimens and has revealed structural information from all regions of the bags in much greater detail than has been available previously. 

All the extracted IOLs were clear in macroscopic and phase appearance with little evidence of viable cells. This suggests that the regression of cells from the anterior surface, seen clinically soon after surgery, ¹² continues with time. The physical presence of the IOL, however, forms a formidable barrier to cell movement within the bag. Cells accumulate and elongate parallel to the interface, as shown by the nuclei and cytoskeletons (Fig. 4) . This barrier effect retards rather than prevents cell coverage of the PC. Trials of experimental designs of IOL (reviewed by Apple et al. ³ ) and alternative materials, ³² intended to enhance this effect, have not been fully effective, although one recent study has shown an improved level of regression of cells from the central PC. ¹⁶  

A striking finding was that in every specimen, even up to 13 years from surgery, the inner surface of the AC was populated by a monolayer of epithelial cells, from the border of the optic to the equator. The phase and TEM appearance and the cytoskeletal and nuclear structure of these cells were indistinguishable from those of a fresh, normal lens epithelium (Fig. 2) . These data indicate the extremely efficient wound-healing response of the anterior cells, because cataract surgical techniques are developed to maximize removal of cells from the AC. Not only are the cleared regions repopulated with identical cells but the repair is remarkably stable. 

Where the two capsular leaflets were in close apposition (Fig. 1B , region 1), two layers of epithelial cells were present. Separation of the capsular leaflets disrupted the apical surfaces of the cells, with loss of the apical actin cytoskeleton. This strongly suggests that such regions result in a closed microenvironment in the bag, between these cells and the equator. Such a closed system certainly develops in the human capsular bag model, in which the formation of the cell layers can be followed with clarity. ^{29 30 33} Within this bounded microenvironment, the anterior lens epithelium is remarkably stable, possibly because of the ability of human LECs to produce autocrine factors to sustain their own growth and development. Cultured capsular bags, ²⁹ grown in a defined medium, ³⁰ have been shown to produce basic fibroblast growth factor, hepatocyte growth factor, ³⁴ transferrin, ³⁵ and transforming growth factor-β (Michael Wormstone, personal communication, June 1999)—all factors affecting cell survival, proliferation, differentiation, and migration. ^{35 36 37}  

The anterior epithelium within the closed system may also regulate the stability of the epithelial cells on the PC and sustain differentiation within the region that formerly made up the bow or equator of the mature lens, (Fig. 1B , region 1a). These cells appear to contain all the elements of differentiating fiber cells, including elongating and degenerating nuclei, ball and socket joints, gap junctions, and expression of MIP26, which is a marker for lens fiber differentiation. ³⁸ Continued differentiation here could be influenced by a fibroblast growth factor gradient at the junction of vitreous and aqueous humors. ³⁶  

Regions adjacent to the IOL, including the central PC and surfaces of the AC flap (Fig. 1B , regions 2–5) remain as an open system and are more likely to be influenced by external factors, such as cytokines in the ocular fluids. ³⁹ These regions contain the light-scattering elements that cause visual disturbance, including cell aggregates formed by regression and multilayering (Figs. 1 8) , three-dimensional arrays of strands of elongated cells (Fig. 6) , and deep, ECM-filled wrinkles (Figs. 6 7) . This study has revealed that the common characteristic of all these cells in the open system is that they had undergone epithelial–mesenchymal transition to fibroblastic–myofibroblastic cell types. They were characterized by the expression of α-sma and by being surrounded by large amounts of ECM, which separated them from the original capsule. In some regions, this ECM contained distinct fibrillar layers characteristic of basement membrane. ^{17 40}  

In every specimen, concentric strands of elongated, α-sma–positive cells were found around the edges of the capsulorhexis, on both posterior and anterior faces of the AC lying on the optic. Although cells have previously been observed on the anterior face, ^{17 41} the concentric alignment of cells into strands surrounding the rhexis does not appear to have been described before. Contraction of these cells, on both AC faces, would produce an effect similar to that of a sphincter muscle and would either pull the rhexis tightly against the IOL if the overlap were good or, if the rhexis were close to the IOL edge, could be the cause of retraction of the rhexis off the IOL. The latter effect is seen clinically and can lead to decentering of the IOL. ^{3 16 42 43}  

This study has revealed the extent of the viable cell populations in capsular bags many years after surgery. While a viable epithelial population exists, even if remote from the central PC, any disturbance of status in the bag could lead to a fresh wave of cell proliferation. Indeed, there have been reports of massive cell proliferation after laser treatment, ⁴⁴ even leading to closure of the posterior capsulotomy. ⁴⁵  

The nature of the various states of cell differentiation observed in these in vivo capsular bags raises certain questions of a fundamental nature. Why does the repopulation pattern on the inner face of the AC that is facing the PC reproduce precisely the original pattern, and yet epithelial cell movement to the region adjacent to the optic and to the outer faces inevitably leads to transdifferentiation? Similarly, why does colonization of the posterior capsule result in a stable epithelial population, provided the cells remain near cells of the anterior epithelium, whereas migration to the central PC is followed by regression and transdifferentiation? Since there is an obvious barrier effect of the IOL on cell distribution, how can this be optimized to reduce PCO? 

Certainly, autocrine control from the anterior epithelial cells must play a major role in maintaining the stability of cell populations, along with the nature of the extracellular matrix. In addition, interaction with factors in the ocular fluids is likely to have a role in the open part of the capsular bag, and this is where transdifferentiation of cells occurs. The mechanisms underlying the subtle differences that determine cell behavior are beginning to be elucidated by observing very long-term cultures of capsular bags in protein-free media in vitro (Michael Wormstone, unpublished data, 1999). This model makes it possible to measure autocrine production and also to determine how added individual growth factors and inhibitors affect cell behavior, in relation to an in vivo endpoint, as detailed in the present study. 

 Supported by The Humane Research Trust, United Kingdom; School of Biological Sciences Research Fund, University of East Anglia, and National Institutes of Health Grant EY-10558.

 Presented in part at the XXVIII Cambridge Ophthalmological Symposium, Cambridge, UK, September 1998.

 Submitted for publication August 17, 1999; revised October 26, 1999; accepted November 16, 1999.

 Commercial relationships policy: N.

 

Figure 1.

 

(A) General appearance of a capsular bag with implanted IOL (donor age 73). Note peripheral light-scattering areas (arrows) and clear central PC. Boxed numbers refer to areas examined in greater detail. (B) Structure of a capsular bag in sectional view. The boxed numbers represent: 1, areas of the bag where the AC and PC are in close apposition; 1a, areas with a Soemmerring’s ring; 2, the rhexis and AC overlying the IOL optic; 3, the PC beneath the optic, with major wrinkles; 4 and 5, PC regions with a scattered cell population, but without major wrinkling. (C through F) Dark-field images of intact capsular bags showing the main aspects of PCO: fibrosis (C, E, and F, arrowheads), Soemmerring’s ring (D, E and F, arrows), and Elschnig’s pearls (E, asterisk). Bars, 2 mm.

Figure 2.

 

Representative micrographs from region 1 of a specimen with sparse, interposed, fiberlike material between the AC (ac) and PC (pc) (A through D, I, J) and a control epithelium (E through H, K). F-actin distribution in the AC cells: (A) cell borders at the apical surface; (B) the dense apical mat and cell borders; (C) cortical actin at the lateral borders of the midcell region; and (D) actin whorls in the basal layer. (E through H) Comparable images of normal epithelial cells from a control lens. (I) TEM of region 1. Confluent cuboid cells on the AC and confluent but flatter cells on the PC. Asterisk, intercellular spaces in PC layer. Arrowheads, close apposition of apical surfaces. (J) DAPI-stained nuclei of cells in region 1; two layers of nuclei in slightly different planes of focus; and (K) comparable nuclei of a control epithelium. Bars, (A, E) 100 μm; (B, C, D, F, G, H, J, and K) 10 μm.

Figure 3.

 

Structure of the equatorial region 1a. (A, B) TEM of a bag with a thin Soemmerring’s ring; AC (ac), nucleated cuboid epithelial cells (ep), fiberlike cells (f) with homogeneous cytoplasm, and very little intercellular space; on the PC (∗), nucleated epithelial lens bow cells (lb) similar to those of the normal lens bow and fiberlike cells (f). (C) Elongated fiberlike cells (f) and nucleated lens bow cells (lb) from the equator-like region of a bag with a thick Soemmerring’s ring. (D) DAPI-stained nuclei of a thick segment of dissected Soemmerring’s ring, close to the equator. Round nuclei of the AC epithelial cells, including short runs of meridional rowlike nuclei (between arrows) over the ovoid nuclei, in several planes of focus, of fiberlike cells. (E, F) Confocal images, at 4-μm separation, of the submembranous F-actin cytoskeleton of the cells of a thicker Soemmerring’s ring. Arrows in (E) are contact points between the basal surface of the cuboid epithelial cells and the fiber ends; open arrows in (F) are parallel membranes of fiberlike cells. (G) TEM of well-ordered fiberlike cells show gap junctions (gj) in the membranes (arrowheads) and interdigitations between cells (arrow). m, mitochondria. (H) Confocal image of fiberlike cells stained with MIP26 antibody to show the presence of the fiber-specific protein in the membranes. Bars, (A, B, C) 5 μm; (D) 100 μm; (E, F, H) 25 μm; (G) 1μ m.

Figure 4.

 

Cell structure close to the IOL haptic at the equator of a capsular bag, corresponding to the area at the extreme left of Figure 1B . (A) Very bright actin staining (arrows), showing close packing of many cells, (the haptic was not removed and appears dark). The outlines of normal anterior epithelial cells (cf. Fig. 2A ) can also be observed. The equator of the bag is at the top arrow of the micrograph. (B) Nuclear staining shows a single layer of elongated nuclei on the equatorial side of the haptic, but multilayers of elongated nuclei on the other side (between arrows). The ovoid nuclei at the bottom of the micrograph indicate the presence of only two layers of epithelial cells. Bar, 100μ m.

Figure 5.

 

Cell structure at the rhexis and on the AC flap over the IOL (Fig. 1 , region 2). (A, B) Elongated cells on the anterior surface of the AC, in bands aligned parallel to the rhexis (arrows) were positive for α-sma (A) and had elongated nuclei (B). Cells on the PC and the posterior face of the AC are out of focus. (C) SEM of well-spread epithelial cells on the anterior face, remote from the rhexis. α-Sma expression (D) was variable, including pronounced stress fibers (arrowheads) in some cells. (E) Phase image of cells on the posterior surface of the AC. Concentric bands of elongated cells close to the rhexis edge (open arrows), with broader patches of cells closer to the rim of the optic. (F) The cells were α-sma positive. Bars, (A, B, E, and F) 100 μm; (C, D) 20 μm.

Figure 6.

 

Images representing structure in the wrinkled regions of the PC beneath the IOL optic (Fig. 1 , region 3). (A) Dark-field micrograph of wrinkles in a capsular bag 2 years after IOL implantation. Light-scattering wrinkles in parallel with the rhexis (arrowheads) and radiating across the PC (arrows). (B) SEM of multilayered strands of cells and areas of bare capsule (pc). Note the absence of cytoplasmic extensions from the cells. (C) TEM of cell strands surrounded by layers of loosely organized ECM (arrows) that separated them from the capsule (∗). (D through G) F-actin (D), α-sma (E, F), and DAPI (G) staining of multilayered cell strands showing elongated morphology and nuclei (white arrows) or intermediate morphology with ovoid nuclei (black arrowheads). Individual cell outlines can be seen in the area between the three arrowheads in (D). Bar, (A) 500 μm; (B, D through G) 50 μm; (C) 10 μm.

Figure 7.

 

Images of the PC wrinkles. (A) Strands of α-sma–positive cells detached from the PC, above wrinkles (open arrows) devoid of a cell population. (B) TEM of a wrinkle containing a group of cells and loose ECM (∗). (C) Two groups of cells embedded in loose ECM, which formed distinct layers close to the cells (arrowheads). One cell group (arrows) showed degenerative changes. (D) A deep PC wrinkle filled with loose ECM and cell debris (arrowheads). Bar, (A) 100 μm; (B through D) 5 μm.

Figure 8.

 

Cells on the PC, in regions without major wrinkles (Fig. 1 . regions 4 and 5). (A) Phase image of multilayered islands of cells on bare PC (∗). Light scatter occurs at the interfaces. (B) Cytoplasmic distribution of α-sma. One large cell (arrow) contained strong stress fibers typical of substrate-attached cells. (C) SEM of groups of cells close to the rhexis with covering of epithelial-like cells (arrows). Bar, (A, B) 100 μm; (C) 50 μm.

The authors thank Peter Davies (ophthalmic surgeon) and Pamela Keeley of the East Anglian Eye Bank for the supply of capsular bags and Sheila Davies for photographic support in the United Kingdom; Ben Willekens and Anneke de Wolf for technical support, and Ton Put, Nico Bakker, and Marina Danzmann for photographic support in The Netherlands; and Michael Wormstone for stimulating discussions.